GPS/GLONASS Combined Positioning Technology And Its Application In Deformation Monitoring

Since the beginning of the21st century, the increasingly intense competition in science, technology and navigation market has brought new developments for the Global Navigation Satellite System (GNSS, Global Navigation Satellite System):the restoring of GLONASS, which takes the lead to turn over a new chapter that not only GPS is available, and rapid construction of other systems such as COMPASS、Galileo etc. also greatly enrich the satellite resources at the meantime. For users, the almost multiplied number and wider coverage of visible satellites, will doubtlessly improve the accuracy, efficiency and reliability of positioning. Therefore, issues such as how to combine different GNSS systems, how to refine models of positioning and weighting based on better knowledge on the systems’differences, how to make full use of multi-GNSS system in fields where traditional way is limited, e.g. deformation monitoring, become research focuses again in the industry.Considering the above reasons, taking combination of GPS and GLONASS as example, the author studied the theories and technologies of combination; from the point of view of the model error processing, discussed the stochastic model and function model error processing, refined them in combined positioning; and then investigated performance of combined system in deformation monitoring, as validating its advantages compare to single GNSS system.The main contributions include:(1)According to the differences between GPS and GLONASS, the combined positioning mathematical model was summed up, formulas of both absolute and relative positioning, and processes of combination were given; accuracies of broadcast ephemeris were analyzed; simple combination of GPS and GLONASS in positioning using broadcast ephemeris was accomplished.(2) Reasonable determination of weights is the key factor in giving full play to advantages of combined system. GPS and GLONASS are heteroscedastic due to their different levels in both design and realization. Based on the thought to combine both priori and posteriori information, using a more precise and comprehensive priori weight matrix, posteriori means like Helmert Variance Components Estimation (VCE) and robust estimation were employed. As refining the stochastic model, an initial weighting solution for combination was recommended.(3) An accurate function model is the prerequisite to getting high-accuracy combined results, due to diversity and complexity of error sources in GNSS fields, function model error may be inevitable and it will be more significant when combine different GNSS systems, its error will seriously affect the weighting and positioning result by making the least square parameter estimates and Helmert VCE deviating from their basic assumption:observation residuals should follow a zero mean normal distribution. Therefore, with a thought that each system’s residuals should tend to be zero mean normal distributed, the author tried to adjust the mathematical model in relative positioning by using semi-parametric estimate, as discussing systematic error processing for combined positioning.(4) Requirements for multi-GNSS data processing software were analyzed, using C++, a core function library that includes GPS, GLONASS and COMPASS was developed, which potentially support the coming Galileo system.(5) Investigated the performance of GPS/GLONASS combined positioning technology in deformation monitoring fields, as verifying its advantages compared to single GNSS system.